This message was deleted.
# harvester
a
This message was deleted.
b
I haven't had that kind of thing happen with Harvester clusters, but I have with other downstream clusters. It was DHCP giving out new addresses and the nodes not having the IPs they started off with.
your 10m comment night show a clue? Anything in journald from those times? (even things that aren't errors)
Lastly, have you run a network layer test?